Abstract 4084: Exploring predictive biomarkers of in vitro sensitivity to azacitidine in breast cancer

Abstract

Abstract The cytidine nucleoside analog azacitidine (AZA) is approved for the treatment of patients with myelodysplastic syndromes and acute myeloid leukemia. In contrast to its activity in patients with these blood cancers, AZA has demonstrated low response rates in clinical trials on unselected patient populations with solid tumors. To realize the potential of AZA in solid tumor patients, we must understand the mechanisms of response and resistance to AZA and develop predictive biomarkers for patient selection. In this study, we sought to identify molecular markers of in vitro sensitivity to azacitidine in breast cancer, using in vitro sensitivity screening of a panel of breast cancer cell lines and gene expression profiling. In vitro sensitivity (EC50) to AZA was determined in 4-day cell viability assays for 21 luminal breast cancer cell lines, while baseline gene expression profiles of cell lines were generated from cells cultured in drug-free growth media. The panel of 21 cell lines displayed markedly differential sensitivity (varied EC50 values) to AZA, revealing both AZA-sensitive and AZA-resistant cells. Analysis of pharmacodynamic (PD) markers for AZA activity revealed substantial reduction of DNA methylation and induction of DNA damage and apoptosis upon AZA treatment of sensitive cell lines in comparison to resistant cell lines. Genes over or under-expressed in association with sensitivity or resistance were identified using two approaches: 1) PAMR analysis, based on fold change across sensitive versus resistant groups (defined by the panel's median EC50) and 2) correlation analysis of baseline gene expression versus AZA EC50. Comparative gene expression analysis of sensitive versus resistant cell lines revealed differential baseline expression of a set of genes involved in multi-drug resistance, cell proliferation and apoptosis. Differential baseline expression of candidate genes in sensitive versus resistant cell lines, as well as their expression after AZA treatment, was validated using QuantiGene mRNA quantification assay. Multi-drug resistance genes were expressed at higher baseline levels in resistant cell lines and showed sustained high expression even after AZA treatment, consistent with a role in AZA resistance. A subset of genes, previously reported as putative tumor suppressors that are silenced by DNA methylation in breast cancer, were found to be under-expressed in sensitive cell lines. AZA treatment resulted in DNA hypomethylation and subsequent re-expression of such genes indicating their predictive value and plausible role in mediating AZA-sensitivity. Taken together, our data provide insights into possible molecular mechanisms and corresponding predictive biomarkers mediating sensitivity to AZA. Further validation of these candidate predictive biomarkers may provide the basis for a hypothesis driven clinical study of AZA in breast cancer or other solid tumors. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 4084. doi:1538-7445.AM2012-4084